Palo Alto Networks to Acquire Koi Security to Advance Agentic Endpoint Protection

Palo Alto Networks is set to bolster its AI security arsenal by acquiring Koi Security, a startup specializing in protecting endpoints from the risks posed by advanced AI agents.

Announced on February 17, 2026, this definitive agreement targets a growing threat: AI agents and tools that wield broad permissions and data access while evading conventional security measures.

These “agentic” systems represent a new frontier in endpoint vulnerabilities, demanding innovative defenses.

Traditional endpoint security focuses on spotting malware through signatures or behavioral anomalies. However, AI agents change the game.

They actively read files, execute scripts, and transfer data across systems with minimal oversight.

Operating as trusted insiders, these tools access sensitive corporate information think customer records or intellectual property, without triggering alerts from tools like antivirus software or endpoint detection and response (EDR) platforms.

Industry observers dub this the “Agentic Endpoint,” an unmanaged attack surface lurking on every laptop, server, and virtual machine.

Attackers are already capitalizing on this blind spot. They exploit flaws in agent frameworks via authentication bypasses, where stolen tokens grant unauthorized entry, or API-driven remote code execution (RCE), allowing malicious code to run undetected.

Techniques include spoofing agent identities to mimic legitimate automation, hijacking credentials for lateral movement, and injecting payloads through extensions, plugins, packages, scripts, or even AI model artifacts.

Unlike classic malware confined to executables, these threats blend into everyday developer workflows, rendering legacy controls obsolete.

Bridging the Agentic Security Gap

Palo Alto Networks plans to weave Koi Security’s technology into its Prisma AIRS platform, which already secures AI-driven operations, and enhance Cortex XDR for broader endpoint visibility.

This integration promises real-time monitoring of AI agent behaviors, automated policy enforcement to restrict risky actions, and proactive malware prevention tailored to agentic environments.

Organizations will gain dashboards showing agent permissions, data flows, and anomaly detection specific to AI tools, closing the gap between development speed and security rigor.

Lee Klarich, Palo Alto’s Chief Product and Technology Officer, described AI agents as “ultimate insiders” with unfettered access yet no traditional guardrails. “This acquisition equips customers with the visibility and control to deploy agentic tools safely,” he stated in the announcement.

Koi CEO Amit Assaraf echoed this, noting that conventional solutions overlook agentic setups. Partnering with Palo Alto will scale Koi’s innovations to enterprise levels, protecting against exploits in tools like those from OpenAI, Anthropic, or custom frameworks.

The deal underscores a shift in cybersecurity: as AI automates tasks from coding to customer service, endpoints evolve into dynamic ecosystems needing agent-aware defenses.

For security teams, this means prioritizing runtime protection over static scans. While financial terms remain undisclosed, the acquisition, expected to close soon, positions Palo Alto to lead in securing the AI era’s hidden risks.

Follow us on Google News , LinkedIn and X to Get More Instant Updates. Set Cyberpress as a Preferred Source in Google

The post Palo Alto Networks to Acquire Koi Security to Advance Agentic Endpoint Protection appeared first on Cyber Security News.